RECENT DEVELOPMENTS IN THE KINETICS AND MECHANISM OF THE DIRECT REDUCTION OF
HEMATITE WITH SOLID CARBON Y. K. RAO
Associate Professor, Henry Krumb School of Mines, Columbia University, New York City, N. Y. 10027 U. S. A.
Abstract : The rate of reaction of hematite fines m) with carbon particles was investigated in the temperature range 850°C to 1087°C. The experimental method consisted of determining the weight loss sustained by a sample of the mixture when maintained at constant temperature under argon atmosphere.
The respective effects of carbon particle size, the hematite to carbon proportion in the mixture and the addition of catalytic reagents were investigated. The reaction occurs by a two-stage mechanistic scheme consisting of (a) reduction of iron oxides by carbon monoxide and (b) oxidation of carbon by carbon dioxide, in other words, carbon passes through the intermediate stage of carbon monoxide of rmation prior to reacting with iron oxides. A chemical kinetic model developed on the basis of carbon solution- loss reaction, viz., stage (b), as rate-controlling represented the results reasonably well. Lithium oxide catalyst was found to enhance the rate of reaction; the addition of ferrous sulfide inhibited the reaction.
These observations are consistent with the proposed mechanism because the carbon solution-loss reaction is known to be promoted by alkali metal salts and inhibited by sulfur-bearing compounds.
INTRODUCTION
That metallic iron can be produced by heat- ing a mixture of hematite fines and carbon parti- cles to a moderately high temperature 850°C), under an inert gas blanket, is well established.
The mechanism of reduction is not fully under- stood despite the numerous earlier studies on the subject1-10. According to the "direct reduction"
mechanism, carbon is presumed to react directly with iron oxide, producing CO along with a certain amount of CO2. Thus:
FeaOb (s) C (s) = Fea0b-I (s) -I- CO (g) 2FeaOb (s) C (s) = 2Fea0b-i (s) + CO2 (g) where a = 1, 2, or 3
when b = 1, 3, or 4
The "direct reduction" is considered to begin at the points of contact between iron oxide and carbon particles. The product of reaction, viz.,
solid iron, forms a shell around the still unreact- ed oxide core. Further conversion of oxide takes place by the diffusion of carbon through the shell to the iron/oxide interface. The interfacial re- action which consumes carbon and releases CO and CO2 is usually assumed to occur rapidly so that the diffusion process becomes the rate-con- trolling step. Indeed the theories proposed by Baukloh and Durrerl and Yun7 maintain that diffusion through iron shell is the critical process in the overall "direct reduction" process.
The gaseous products of reaction, i.e., CO and CO2, if not removed from the reaction system as soon as these are produced tend to participate in the reduction process as outlined below :
Fea0b (s) + CO (g) = FeaOb_1 (s) CO (g) (1) C (s) + CO2 (g) = 2C0 (g) (2)
1.0 YUN 0961)
BA11KLOH ET AL 11930
2 0.8 0.26
0.24
0.026 0.024 o 0.020 0.20
cc 0.6 4 z 2 I-
U 0.016
0.16
„r,_, 0.12 0.012 cc 4 0.4
cc
O.oe 0.005 vl 0004 3
0.04
til•,1 0
o0 4 6 12 16
TIME HOURS 20 • 140 200 240 260 320 360 400
TIME, MINUTES 1000•C
900° -
*XV .4.4
`a' and `b' have the same values as specified earlier. Arkharov et al6, Baukloh and Durrerl.
and Yun7 performed direct reduction experiments under vacuum to eliminate the effects of CO and CO2. In Figure 1, the direct reduction kinetics determined by different investigators are shown, The results are plotted according to Jander's equational, reproduced..below:
[ 1 —(1 f ) "3 12 = r.
(3)
where 'r' is the initial radius of the oxide particle and 'f' is the fraction 'reacted at time Eqn. (3) was derived on the assumption that the reduc- tion is controlled by the rate of diffusion through the product layer. Athough the results of Baukloh and Durrerl appear to be in good agreement with the predicted linear behavior, concern. has been voiced in the literature3,4,18 about the accuracy of these authors' experimental method. Furthermore, working with more string.
ent experimental conditions, particularly a high vacuum of 5 x 10-4 mm, Yun7 obtained direct reduction kinetic data which, as shown in Figure 1, deviates considerably from the predicted linear behavior. The disagreement between these two studies suggests that secondary reduction effects produced by CO and CO2 might have overshadow- ed the "true" direct reduction kinetic effects.
The present study was begun with the aim of investigating the kinetics of the overall reduc- tion process—not only the solid/solid reaction but also the gas/solid interactions—in the hematite-F carbon mixture -heated to a high temperature. In- sight into the mechanism of the reduction pro- cess was gained by investigating the effects of catalytic additives.
EXPERIMENTAL
The materials used in this study were Baker reagent grade hematite powder (99.5 ± wt.%i) and high purity carbon, the latter supplied by Ultra Pure Carbon Co., Bay City, Michigan. To determine the effect of the particle size on reduction kinetics three different size- fractions of carbon were employed: a fine-sized fraction (-325 mesh), a medium-sized fraction (-100 to + 115 mesh) and a coarse-sized fraction (-48 to + 60 mesh). The average hematite par- ticle size was fdund to be about 1 p,m; only one size-fraction was employed in this work.
The samples were prepared by mixing he- matite and carbon powders in the designated pro- portion. Samples weighing about 1 g. were taken from the prepared mixtures for carrying out the weight-loss measurements. Pelletized samples, 5/16" x 1/2", were also used in some experiments.
The details of the experimental procedure are omitted here as a clear account of it has been published earlier17. The weight-loss sustained by the sample when held at constant temperature in an argon 'atmosphere, was determined at several time intervals. Based on these weight- loss measurements, the "fractional reaction" values were computed using the following relationship.
f = ( 100 AWt/ NW° ) (4)
where (100 Wt/Wo ) is the weight loss per- cent at time `V and N is the weight loss per- cent corresponding to complete ,reduction. N was determined experimentally for each mixture;
and N = 38.7, 37.0, 39.5 and 30.8, respectively, for mixtures Fe203 + 3C, Fe2O3 + 1.5C, Fe2O3 -t- 2.25C and Fe2O3 + 9C.
RESULTS
The experimental data is presented in the form of "fractional reaction" vs "time" plots in Figure 2 through 6.
FI6.1 DIRECT' RC')UCTION OF Fs203 BY CARBON; RESULTS PLOTTED ACM/ROMS TO
WINDER'S EQUATION FIG. 2 THE` EFFECT OF tEMPERATURE.ON THE REDUCTION OF HEMATITE 8Y CARBON
GO' 50
0 20 30 40
TIME, MINUTES
OF TEMPERATURE ON THE REDUCTION -OF HEMATITE
F10: 3 THE EFFECT BY CARBON
0,0
0.60
0.40
0.20
120 160
0 40 so
TIME, MINUTES
20 40 60 80 120 140
TIME. MINUTES I.0
0.8
O
0 0.6
.0.4
0.2
0
FIG.4 THE EFFECT OF PARTICLE SIZE OF CARBON ON THE REDUCTION OF HEMATITE
tures having different Fee 03 /C ratios are given in Figure 5. The rate promotion caused by the addition of 5% by wt. Li20 and the rate inhibi- tion resulting from the presence of 5% by wt.
FeS are shown in Figure 6.
DISCUSSION:
As no attempt was made in this study to withdraw the gaseous products from the site of reaction, it is surmised that both CO and CO2 fully partake in additional reactions as given by eqns. (1) and (2). It has been proposed 5,9,10,17 that reactions (1) and (2) together account for the entire reduction occurring in a mixture of iron oxide and carbon heated to an elevated tem- perature. During the initial stage of the reduc- tion process, CO is formed by one or more of the following mechanisms: (1) Oxygen of the air en- trapped in the porous sample combines with car- bon (2) Oxygen chemisorbed on carbon surfaces liberated as CO (3) Carbon reacts with oxygen released by the dissociation of iron oxides and (4) direct reduction occurring at the points of contact between carbon and oxide particles.
Carbon monoxide thus generated attacks he- matite particles and reduces them to lower oxides and metallic iron.
3Fe2O3 (s) + CO (g) = 2Fe304 (s) CO2 (g) (5) Fe304 (s) CO (g) = 3Fe0 ( s ) CO2 (g) (6) FeO ( s ) CO (g) = Fe ( s ) CO, (g) (7)
FRACTIONAL REACTION, f
sv. 5 THE EFFECT OF THE "Ft203/C° RATIO OF THE MIXTURE ON THE REDUCTION OF HEMATITE
The effect of temperature on the reduction kinetics is shown in Figures 2 and 3; only the fine-sized carbon was used. In Figure 4, the in- fluence of carbon particle size is demonstrated.
The proportion of Fe203 to C was maintained the same at 1 to 3 in the foregoing experiments.
The results of experiments performed with mix-
1.00
0.80 o •
0.60 cc , 41 0.40 0 U LI 0.20
0
1.00
g 0.80
U
tcl 0.60 I
-J
a z 0.40 0 17- a 4 0 20
Ii.
0
FIG.6 THE EFFECT OF PROMOTIVE (L120) AND INHI8ITIVE (F.S) AGENTS ON' THE REDUCTION OF•HEMATITE
FRACTIONAL REACTI
1 dWc
Wc dt t wic
poo
kAik Pc021720 kA LCl P002
, , ilia of C - min
Incidentally, eqns. (5) through (7) can be gene- rated from eqn. (1) by substituting appropriate values for the suffixes 'a' and 'la' respectively.
The gaseous product of reduction, i e., CO2 reacts with carbon particles and produces carbon monoxide, according to eqn. (2). This is the sa-called solu- tion-loss or Boudouard reaction. Since the rate of consumption of CO gas, by reactions (5) through (7), is much faster than its regeneration by reac- tion (2), the gas-phase within the porous sample becomes progressively less reducing until a steady- state is reached; whereupon the overall reduction
process occurs at much the same rate as the solution-loss reaction. There have been a num- ber of investigationsi2-16 to elucidate the mecha- nism of the reaction between carbon and carbon dio- xide. The overall reaction is considered to occur by a sequence comprised of two elementary reactions.
1. Reversible oxygern-exchange reaction involv- ing transfer of oxygen from CO2 to carbon surface.
k A
CO2 (g) " c0 (g) (0) c
k'
(2.A)where pc and (0) , denote a "free reaction site"
and a "chemisorbed oxygen atom" on the carbon surface; and
2. The chemisorbed oxygen combines with a carbon atom which is -liberated as carbon monoxide.
kg
CO (9) + P
c (2.B) The rate equation for reaction (2) can be written asi6.ers20 and Turkdogan and Vinters2l, have establish- ed that iron, in amounts as small as a few parts per million, greatly enhances the rate of reaction between carbon and carbon dioxide. The rate promotion by the iron catalyst may be attributed to a substantial increase in the number of reac- tion sites, i.e., z , near the iron/ carbon contact points. An increase in z , according to eqn.
(8), results in an enhanced rate of reaction. Much of the catalytic activity of iron, however, is des- troyed when it alloys with carbon or forms vvus tite.
CHEMICAL KINETIC MODEL :
The kinetic data was interpreted in the light of the two-stage reaction mechanism discussed earlier. The reduction of iron oxides by CO, i.e., eqns. (5)" through (7), occurs rapidly so that , one may assume the gas-phase within the porous sample to be in virtual equilibrium` with "FeO" and Fe. As the penetration of argon gas is bound to be limited to the superficial layers of the sample, the gas- phase may be assumed to consist of CO and CO2 only. The values of the equilibrium P CO2 /P CO ratio for reaction (7) are listed in Table 122. The principal resistance to the overall reduction pro- cess stems from the chemical kinetics of the solution-loss reaction which constitutes the chief mode of CO generation within the sample.
Let n° CO2 and n °CO denote the number of moles of CO2 and CO within the porous sample at any time 't'. Subsequently, the occurrence of reaction (2) over a time-span of 'cIt' results in the consumption of 'x' g. moles of CO2 and simul- taneous production of '2x' g. moles of CO gas.
Let' 0'. be the fraction of CO gas which is utilized in the iron oxide reduction according to eqn. (7).
2x ( fiFe0(s) P.c9(9) PFe(s) PCO2(9)
(0)c
As a result of these reactions, the number of moles of CO2 and CO within the sample now be- come nCO2 and nCO respectively. By molar balance,
In eqn (8), We denotes the mass of carbon at any time 't', E c is the total number of reaction
sites on carbon in
g.
mole/g, of C., kA, k'A andx + 2 x p kB
are specific rate constants and Pco2, Pcoare the partial pressures of CO2 and CO in the and • • gas-phase.
The possibility of freshly reduced iron, pro- duced by reaction (7), catalyzing the carbon
solution cannot be disregarded. The investiga- Since "near" equilibrium conditions are tions of King and Jones19, Walker and co-work- -.,•assdined to prevail With. respect to reaction (7), (9.A)
n co o
'CO
4.-2 x - 2 x /3 (9.B)
no
CO-2 n
CO2 =
n° °CO
C.0
where Ke (=PCOSPC.0) is the equilibrium stant for reaction (7). From. eqns. (9) and
1
- (1/2 + 2 Ke) Ke
0
T k
K = P /P e CO2 CO
a 1 atm
*
P
CO2 P
co2
1123 0.619 0.382 0.618 0.691
1180 0.551 0,355 0.645 0.678
1220 0.511 0.338 0.662 0.669
1230 0.502 0.334 0.666 0.667
1260 0.487 0.327 0.673 0.664
1.28 0.460 0.315 0.685 0.657
1310 0.438 0.304 0.696 0.652
1360 0.405 0.288 0.712 0.644
**
x
1.352 1.335 1,323 1.321s 1.317 1.309 1.303 1.292
for an isothermal sample of a given volume, as negligible over the temperature range investi- gated here. Thus,
(10) ~
1 dWc
-Va -Tr = 110' min-1 ( (12) where It c is the rate constant defined by eqns.
con- (8) and (12). The total weight loss, LsWe sustain- (10). ed by the mixture over a time-span 't' can be
computed from the rate of carbon solution-loss (11) reaction.
The values of {3 at various temperatures are list- ed in Table I.
The_rate at which carbon gasified is given by eqn. (8). At a given temperature; the Magni- tude of the right side of , this equa#on is deter- mined by (1) The values of the specific rate con- stants kk
kA
and k s and (2)" the composition of the gas-phase, i.e.,Pco
and P co for the hetergeneous equilibrium "FeO"/Ile/CO CO, To a good approximation the variation in the Pco2 and Pco values (Table 1) may be regardedEquation (12) is integrated using the bound- ary conditions W c = c at t= 0 and W C = W g at t t . This gives
We = Wc° exp. ( Rct ) (13) The amount S of carbon gasified during time-span
`t' is
&WC = W"c Wc = woo [ 1 - exp. ( Rct )1 (14) As per the stoichiometry of reaction" (2), for 12 gms. of carbon gasified, 2 g: moles of CO are
TABLE 1 : Values of utilization factor
13 = 1-(1/2+2 Ke)
** A = 0, 475 (1 + 2.67 0 ), for Fe
203 + 3C mixture 68
Published data 23
produced; of which 24 moles are utilized in iron oxide reduction. In other words, 2f3g.
atoms of oxygen are extracted. Thus the weight loss sustained by oxide in the sample is given by
32 w
12 t " " c
The total weight loss during time-span ‘t, is
A Wt = W
c (1 + 2.67 /3)
(16) From eqns. (14) and (16)As
W t Wcs (1 + c 2.67 fl) [1-exp (-R t) I (17)
For Fe201 + 3C. mixture, W°c = (36/195.7) Wo, where Wo is the initial weight of the mixture.
From eqns. (4) and (17):
(36) (1+2 67 3) [1-exp. (-R t)
where
X, = 0.4753 (1 + 2.67
13) (20) The values of 'X' at different temperatures are given in Table 1.The experimental data was recast into "log (X-f)" vs. "t" plots as shown in figures 7 and 8.
The agreement with the model is reasonably good at temperatures ranging from 957 to 1087 deg.
C. At lower temperatures,, i.e. at 907 deg. and 947 deg. C, the plots exhibit three different stages ; each of which occurs at a slower rate than the preceding one. As discussed in an earlier publication17, this behaviour may be attributed to the relative ease with which the iron catalyst gets deactivated (through reoxi- dation) at lower temperatures. Furthermore, the gas-phase composition, which .has been assumed to remain constant over the entire time, might change at the completion of the Fea03-> "FeO"
conversion.
The values of the rate constant "Rc" evalu- ated from the steepest slopes of these plots are listed in Table II. The enthalpy of activation was
A ox = o
(15)f (0.387) (195.7) Rearranging the terms :
log
f) = (RI2.303) t log X
(18)(19)
TABLE II : Values of Rate Constant "R"
0 -1
:T K Rmin
c' k /k
A B k'/k A B
)
A 10/T
-4 4
1123 1.152x10 3.06 1973 5.12x10 8.905 3.9385 3.2910
-4 -3
1180 8.356x10 2.65 531.4 1.13x10 8.475 3.0780 2.948
-3 -3
1220 5.890x10 2.41 227.7 3.65x10 8.197 2.2300 2.438
-3 _3-
1230 6.910x10 2.36 185.8 3.65x10 8.130 2.1600 2.438
-2 -3
:1260 1.565x10 2.21 103.1 4.72x10 7.937 1.8050 2.326
-2 -3
1280 1233;00 2.12 70.5 4.92x10 7.812 1.6510 2.308
-2 . -3
'1310 4.788x10 2.00 40.9 6.58x10 7.634 1.3200 2.182 1360 0.134 2 1.82 17.4 8.99x10 -3 7.353 0.8720 2.046 4 -log R -log (k A
kA C = (R /720 P ) [1 +(k'fk) P
c CO A B
2
( k ) P CO A B CO
2.
The units of k are g. mole/g. of C - atm. - sec.
A C
0 R, k,Z, -1.0
72 I 4 k•oal•/mola -1.5
-2.0
/ AA
23 34 k-031.finole •■
♦ 4
-3.0
- 15
-4.07.2 74
,c55 3 6 I•obtroole
• • 0 ♦ ♦ ♦
♦
110 3 5 k.c01./rnole-7-11
7.6 7.8 8.0 8.2 84 8.6 81.8 9.0 FIGURE Ps RESULTS REPLOTTED ACCORDING TO EQUATION OP)
I, mIrritet F/GUtt 14 RESULTS REPLOTTED ACCORDING TO EQUATION on
determined by construdting a "log 1.2c" versus
"10;000/T" plot, as shown in Figure 9. Within the temperature range 957 deg. to 1087 deg. C the enthalpy of activation was found to be 72÷ 4 k.
cal./mole. Over the lower temperature range, value of 110 + 5 k. cal. / mole. This larger the enthalpy of activation has a much larger value, it must be emphasized, includes the com- bined effects of (1) temperature itself, (2) the cata- lyst action of iron which is markedly influenced by temperature and (3) gas-phase composition which is also a function of temperature. A clear- cut separation between the effect of temperature and the effects of other parameters on the rate constant, R is not possible. As a result, no de- finitive conclusions may be drawn at this stage concerning the derived activation enthalpies.
A better understanding of the reaction mechanism can be gained by reinterpreting the rate constant (R0 ) data in the light of experi- mental data of Wu, reported by Reif23, for kA/kB and k'A/kB The values of "kA c" were calculated at different temperatures using these data and the Pco2, Pco values listed in Table
1. The results are summarized in Table II. The Arrhenius plot, "log kA E ," vs "10,000/T", ex- hibits two distinct regions, as shown in Figure 9. In the high temperature region, i.e., from 947 deg. to 1087 deg. C, the activation enthalpy was found to be 23 + 4 k. cal. I mole. A value of 55 + 6 k. cal. / mole was obtained for the lower temperature region.
In general, a catalyzed reaction is believed to manifest ,a much weaker temperature-depen- dence as compared to the uncatalyzed reaction.
For, the uncatalyzed carbon solution-loss reaction, Mentser and Ergun24 reported a value of 53 k.
cal/mole as the activation energy associated with the forward specific rate constant, "kA E c". By comparing this value with the present data, two conclusions may be drawn: (1) at lower temper- atures the catalysis of solution-loss reaction by iron catalyst is much less significant than previ- ously realized; and the decrease in the rate of the reduction process may simply be due to a change in gas-phase composition. Therefore, the derived activation enthalpy of 55 k. cal./mole, within experimental error, may be regarded as the true activation energy for the forward reaction rate constant, "kA E ". (II) the considerable lowering of the activation enthalpy (almost 30 k. cal./mole) for the high temperature region points out that in this range the iron catalyst
10,000/T
FIGURE 9: AitlifiEN US PLOT
PERCENT ittAofto
actively partakes in the enhancement of the rate of solution-loss reaction.
VALIDITY OF THE PROPOSED MECHANISM:
That carbon solution-loss reaction gov- erns the overall reduction process emerges as the central feature of the proposed mechanism.
Recent studies25,26 from this laboratory have .dealt with the general problem of catalysis of carbon-gas reactions. The results of these studies, giVen in Figure 10, clearly show that the addi-
tibn of 1.5% by wt. Li20 considerably enhances the rate of solution-loss reaction. These obser- vations are in complete accord with the results repOrted here, i.e., Figure 8, on the effect of lithium oxide on the rate of reduction process.
Chufarov and Antonova27 reported that sulfur- bearing compounds tend to retard the rate of carbon solution-loss reaction. The addition of fer- rus sulfide does indeed produce rate-inhibition as indicated in Figure 6.
SUMMARY AND CONCLUSIONS:
(1) The direct reduction of hematite fines by solid carbon occurs via gaseous intermediates CO and CO,. (2) The reduction kinetic data can be meaningfully interpreted by means of a chemi- cal kinetic model, developed on the basis that carbon solutiort-loss reaction controls the over- all rate. (3) The observed effects of Li20 and F'eS are in agreement with the reported effects of these reagents on the carbon solution-loss re- action.
References :
1.. W. Baukloh and R. Durrer: Arch. Eisen- huttenw, 1931,. vol. 4, pp. 455-60.
2. W. Baukloh and G. Zimmerman: Stahl u.
Eisen, 1933, vol. 53, pp. 172-73.
3. A.A. Baikov and A.S. Tumarev: Izv. Akad.
Nauk SSSR, 1937, No. 1, p. 25.
4. O.A. Esin and P.V. Gel'd: Uspekhi Khim., 1949, vol. 18, pp. 65&-81.
5. B. Baldwin: J. Iron and Steel Inst. (Lon- don), 1955, vol. 179, pp. 30-36.
6. V. Arkharov, V. N. Bogoslovskii, M. G.
Zhuravleva, and Q. L Chufarov: Zh. Fiz.
Khim., 1955, vol. 29, No. 2, p. 272.
7. T.S. Yun: Trans. ASM, 1961, vol. 54, pp.
129i-142.
8. E.A. Krasheninnikov and E.P. Timofeev:
Izv. Akad. Nauk SSSR, Metal, 1968, No. 2, pp. 10-17.
9. H.K. Kohl and B. Marincek: Arch. Eisen- huttenw., 1967, vol. 38, pp. 493500.
10. K. Otsuka and D. -Kunii: J. Chem. Eng.
Japan, 1969, vol. 2, pp. 46-50.
11. W. Jander: Z. Anorg. Allgern. Chem., 1927, vol. 163, p. 11; Z. Anorg. Allgem. Chem., 1927, vol. 166, p. 31.
12. P.L. Walker, Jr., F. Rusinko, Jr. and L.G•
Austin: Advan. Catalysis, 1959, vol. 11, pp.
134-321.
13. E.T. Turkdogan and J.V.. Vinters: Carbon, 1969, vol. 7, p. 101. Carbon, vol. 8, pp. 3853.
14. S. Ergun: J. Phys. Chem., 1956, vol. 60, p. 480.
15. H.J. Grabke: Carbon, 1972, vol. 10, pp.
587-599.
16. Y.K. Rao and B.P. Jalan: Met. Trans; 1972, vol. 3, pp. 2465-77.
17. Y.K. Rao: Met. Trans., 1971, vol. 2, pp.
1439-47.
18. P.V. Gel'd, V.G. Vlasov, and N.N. Sere- brennikov: Doklody. Akad. Nauk SSSR, 1951, vol. 78, No. 4, pp. 693-96.
19. J.G. King and J.H. Jones: J. Inst. Fuel, 1931, vol. 5, p. 39.
20. P.L. Walker, Jr. M. Shelef, and R.A.
Anderson: Chemistry and Physics of Car- bon, vol. 4, pp. 287-383, Marcel Dekker, New York, 1968.
21. E.T. Turkdogan and J.V. Vinters: Carbon, 1972,T vol. 10, pp. 97-111.
00 10
TIME, MINUTES
Fl GTO EXPERIMENTAL RESULTS FOR CATALYST-DOPED CARBON PELLET' CATALYST CONCENTRATION= In Li 20
22. 0. Kubaschewski, E. LL. Evans and C.B Alooek ; Metallurgical Thermochemistry,
Pergamon Press, New York, 1967.
23. A.E. Reif: J.. Phys. Chem, 1952, vol. 56, p. 785; J. Phys. Chem., 1952 vol. 56, p. 778.
24. M. Mentser and S. Ergun: Carbon, 1967, vol. 5, p. 331.
25. B.P. Jalan: M.S. Thesis, Mineral Engineer- ing, Columbia University, New York, 1970.
26. P. Mukerjee: M.S. Thesis, Mineral Engi- neering, Columbia University, New York, 1973.
27. G.I. Chufarov and M.F. Antonova: Bull.
Akad. Sci; U.S.S.R; Classe Sci. Tech., 1947, vol. 4, pp. 381-89.
72
SESSION II — SPONGE IRON — FUNDAMENTAL ASPECTS, PRESENT STATUS AND FUTURE DEVELOPMENT
Monday, 19 February, 1973
The Session was convened at 230 P. M. The Chairman was Dr. Samir Taher, Chairman &
Managing Director of Delta Steel Mill Co., Cairo, Egypt, the Co-Chairman was Mr. H. Chr. Andersen, Manager, Metallurgical DepartMent, Elkem-Spigerverket A/S, Oslo, Norway and the Rapporteur was Dr. A. B. Chatterjea, Deputy Director, National Metallurgical Laboratory, Jamshedpur.
